Genetic Algorithms-Based Design of Vibration Suppression Compensator for 2-Mass Resonant System

Abstract

This paper presents an autonomous compensator design using Genetic Algorithms (GA) for vibration suppression control in 2-mass resonant systems. The control system is composed of a robust 2-degrees-of-freedom compensator based on the coprime factorization description. In general, the optimal design for free parameters of the compensator requires the complicated numerical calculation process under given and/or identified plant parameters. In this paper, autonomous algorithms for both the identification of plant parameters and the design of compensator are proposed, where the optimal set of the plant parameters and the free parameters of compensator can be identified in each design phase by the optimization ability of GA. The effectiveness of the proposed optimal design can be verified by experiments paying attention to the vibration suppression characteristics and the robustness against variations of mechanical parameters.